End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +3108 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: stratos-unverified-mix-scaled-0.125
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# stratos-unverified-mix-scaled-0.125
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: stratos-unverified-mix-scaled-0.125
|
|
|
|
| 16 |
|
| 17 |
# stratos-unverified-mix-scaled-0.125
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/stratos-unverified-mix-scaled-0.125 dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 666359318511616.0,
|
| 4 |
+
"train_loss": 0.5066342979125237,
|
| 5 |
+
"train_runtime": 9306.6951,
|
| 6 |
+
"train_samples_per_second": 4.514,
|
| 7 |
+
"train_steps_per_second": 0.047
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 666359318511616.0,
|
| 4 |
+
"train_loss": 0.5066342979125237,
|
| 5 |
+
"train_runtime": 9306.6951,
|
| 6 |
+
"train_samples_per_second": 4.514,
|
| 7 |
+
"train_steps_per_second": 0.047
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 438,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00684931506849315,
|
| 13 |
+
"grad_norm": 5.92502498626709,
|
| 14 |
+
"learning_rate": 2.2727272727272729e-07,
|
| 15 |
+
"loss": 0.8592,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0136986301369863,
|
| 20 |
+
"grad_norm": 5.975358009338379,
|
| 21 |
+
"learning_rate": 4.5454545454545457e-07,
|
| 22 |
+
"loss": 0.8441,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02054794520547945,
|
| 27 |
+
"grad_norm": 6.038137435913086,
|
| 28 |
+
"learning_rate": 6.818181818181818e-07,
|
| 29 |
+
"loss": 0.8847,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0273972602739726,
|
| 34 |
+
"grad_norm": 6.023561000823975,
|
| 35 |
+
"learning_rate": 9.090909090909091e-07,
|
| 36 |
+
"loss": 0.8802,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03424657534246575,
|
| 41 |
+
"grad_norm": 5.958974361419678,
|
| 42 |
+
"learning_rate": 1.1363636363636364e-06,
|
| 43 |
+
"loss": 0.9015,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0410958904109589,
|
| 48 |
+
"grad_norm": 5.522011756896973,
|
| 49 |
+
"learning_rate": 1.3636363636363636e-06,
|
| 50 |
+
"loss": 0.8771,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04794520547945205,
|
| 55 |
+
"grad_norm": 5.2923994064331055,
|
| 56 |
+
"learning_rate": 1.590909090909091e-06,
|
| 57 |
+
"loss": 0.8668,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0547945205479452,
|
| 62 |
+
"grad_norm": 4.468110084533691,
|
| 63 |
+
"learning_rate": 1.8181818181818183e-06,
|
| 64 |
+
"loss": 0.8269,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06164383561643835,
|
| 69 |
+
"grad_norm": 3.9916787147521973,
|
| 70 |
+
"learning_rate": 2.0454545454545457e-06,
|
| 71 |
+
"loss": 0.7879,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0684931506849315,
|
| 76 |
+
"grad_norm": 2.3367724418640137,
|
| 77 |
+
"learning_rate": 2.2727272727272728e-06,
|
| 78 |
+
"loss": 0.7631,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07534246575342465,
|
| 83 |
+
"grad_norm": 2.0509274005889893,
|
| 84 |
+
"learning_rate": 2.5e-06,
|
| 85 |
+
"loss": 0.7555,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0821917808219178,
|
| 90 |
+
"grad_norm": 1.9280574321746826,
|
| 91 |
+
"learning_rate": 2.7272727272727272e-06,
|
| 92 |
+
"loss": 0.7413,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.08904109589041095,
|
| 97 |
+
"grad_norm": 1.8343186378479004,
|
| 98 |
+
"learning_rate": 2.954545454545455e-06,
|
| 99 |
+
"loss": 0.7588,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.0958904109589041,
|
| 104 |
+
"grad_norm": 3.5027122497558594,
|
| 105 |
+
"learning_rate": 3.181818181818182e-06,
|
| 106 |
+
"loss": 0.7552,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.10273972602739725,
|
| 111 |
+
"grad_norm": 3.766814947128296,
|
| 112 |
+
"learning_rate": 3.409090909090909e-06,
|
| 113 |
+
"loss": 0.7653,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1095890410958904,
|
| 118 |
+
"grad_norm": 3.9283227920532227,
|
| 119 |
+
"learning_rate": 3.6363636363636366e-06,
|
| 120 |
+
"loss": 0.7895,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.11643835616438356,
|
| 125 |
+
"grad_norm": 3.853430986404419,
|
| 126 |
+
"learning_rate": 3.863636363636364e-06,
|
| 127 |
+
"loss": 0.7636,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1232876712328767,
|
| 132 |
+
"grad_norm": 3.0980968475341797,
|
| 133 |
+
"learning_rate": 4.0909090909090915e-06,
|
| 134 |
+
"loss": 0.7197,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.13013698630136986,
|
| 139 |
+
"grad_norm": 2.2356159687042236,
|
| 140 |
+
"learning_rate": 4.3181818181818185e-06,
|
| 141 |
+
"loss": 0.6567,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.136986301369863,
|
| 146 |
+
"grad_norm": 2.1344797611236572,
|
| 147 |
+
"learning_rate": 4.5454545454545455e-06,
|
| 148 |
+
"loss": 0.7121,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.14383561643835616,
|
| 153 |
+
"grad_norm": 1.604323148727417,
|
| 154 |
+
"learning_rate": 4.772727272727273e-06,
|
| 155 |
+
"loss": 0.7,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1506849315068493,
|
| 160 |
+
"grad_norm": 1.3561691045761108,
|
| 161 |
+
"learning_rate": 5e-06,
|
| 162 |
+
"loss": 0.6897,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.15753424657534246,
|
| 167 |
+
"grad_norm": 1.3645248413085938,
|
| 168 |
+
"learning_rate": 5.2272727272727274e-06,
|
| 169 |
+
"loss": 0.668,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.1643835616438356,
|
| 174 |
+
"grad_norm": 1.4382814168930054,
|
| 175 |
+
"learning_rate": 5.4545454545454545e-06,
|
| 176 |
+
"loss": 0.6585,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.17123287671232876,
|
| 181 |
+
"grad_norm": 1.3407988548278809,
|
| 182 |
+
"learning_rate": 5.681818181818183e-06,
|
| 183 |
+
"loss": 0.6639,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.1780821917808219,
|
| 188 |
+
"grad_norm": 1.243182897567749,
|
| 189 |
+
"learning_rate": 5.90909090909091e-06,
|
| 190 |
+
"loss": 0.6519,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.18493150684931506,
|
| 195 |
+
"grad_norm": 0.9758120775222778,
|
| 196 |
+
"learning_rate": 6.136363636363637e-06,
|
| 197 |
+
"loss": 0.6133,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.1917808219178082,
|
| 202 |
+
"grad_norm": 0.9279192686080933,
|
| 203 |
+
"learning_rate": 6.363636363636364e-06,
|
| 204 |
+
"loss": 0.6404,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.19863013698630136,
|
| 209 |
+
"grad_norm": 1.1322438716888428,
|
| 210 |
+
"learning_rate": 6.590909090909091e-06,
|
| 211 |
+
"loss": 0.6385,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2054794520547945,
|
| 216 |
+
"grad_norm": 1.0254402160644531,
|
| 217 |
+
"learning_rate": 6.818181818181818e-06,
|
| 218 |
+
"loss": 0.6447,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.21232876712328766,
|
| 223 |
+
"grad_norm": 0.7659736275672913,
|
| 224 |
+
"learning_rate": 7.045454545454546e-06,
|
| 225 |
+
"loss": 0.6244,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2191780821917808,
|
| 230 |
+
"grad_norm": 0.8017905354499817,
|
| 231 |
+
"learning_rate": 7.272727272727273e-06,
|
| 232 |
+
"loss": 0.6137,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.22602739726027396,
|
| 237 |
+
"grad_norm": 0.8775509595870972,
|
| 238 |
+
"learning_rate": 7.500000000000001e-06,
|
| 239 |
+
"loss": 0.6211,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2328767123287671,
|
| 244 |
+
"grad_norm": 0.8729256987571716,
|
| 245 |
+
"learning_rate": 7.727272727272727e-06,
|
| 246 |
+
"loss": 0.6034,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.23972602739726026,
|
| 251 |
+
"grad_norm": 0.6592076420783997,
|
| 252 |
+
"learning_rate": 7.954545454545455e-06,
|
| 253 |
+
"loss": 0.6151,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.2465753424657534,
|
| 258 |
+
"grad_norm": 0.5613800883293152,
|
| 259 |
+
"learning_rate": 8.181818181818183e-06,
|
| 260 |
+
"loss": 0.5855,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.2534246575342466,
|
| 265 |
+
"grad_norm": 0.6701193451881409,
|
| 266 |
+
"learning_rate": 8.40909090909091e-06,
|
| 267 |
+
"loss": 0.6001,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2602739726027397,
|
| 272 |
+
"grad_norm": 0.708198070526123,
|
| 273 |
+
"learning_rate": 8.636363636363637e-06,
|
| 274 |
+
"loss": 0.5876,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.2671232876712329,
|
| 279 |
+
"grad_norm": 0.5600025653839111,
|
| 280 |
+
"learning_rate": 8.863636363636365e-06,
|
| 281 |
+
"loss": 0.5883,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.273972602739726,
|
| 286 |
+
"grad_norm": 0.5653926134109497,
|
| 287 |
+
"learning_rate": 9.090909090909091e-06,
|
| 288 |
+
"loss": 0.5614,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.2808219178082192,
|
| 293 |
+
"grad_norm": 0.7606906890869141,
|
| 294 |
+
"learning_rate": 9.318181818181819e-06,
|
| 295 |
+
"loss": 0.6156,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2876712328767123,
|
| 300 |
+
"grad_norm": 0.5844655632972717,
|
| 301 |
+
"learning_rate": 9.545454545454547e-06,
|
| 302 |
+
"loss": 0.586,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2945205479452055,
|
| 307 |
+
"grad_norm": 0.49894917011260986,
|
| 308 |
+
"learning_rate": 9.772727272727273e-06,
|
| 309 |
+
"loss": 0.5747,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3013698630136986,
|
| 314 |
+
"grad_norm": 0.5528267621994019,
|
| 315 |
+
"learning_rate": 1e-05,
|
| 316 |
+
"loss": 0.5419,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.3082191780821918,
|
| 321 |
+
"grad_norm": 0.6437212228775024,
|
| 322 |
+
"learning_rate": 9.999841055681184e-06,
|
| 323 |
+
"loss": 0.568,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.3150684931506849,
|
| 328 |
+
"grad_norm": 0.5419601798057556,
|
| 329 |
+
"learning_rate": 9.999364232830053e-06,
|
| 330 |
+
"loss": 0.5715,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.3219178082191781,
|
| 335 |
+
"grad_norm": 0.4963567554950714,
|
| 336 |
+
"learning_rate": 9.99856956176192e-06,
|
| 337 |
+
"loss": 0.5492,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.3287671232876712,
|
| 342 |
+
"grad_norm": 0.7037913203239441,
|
| 343 |
+
"learning_rate": 9.997457093000165e-06,
|
| 344 |
+
"loss": 0.5849,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3356164383561644,
|
| 349 |
+
"grad_norm": 0.7691269516944885,
|
| 350 |
+
"learning_rate": 9.996026897273024e-06,
|
| 351 |
+
"loss": 0.585,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.3424657534246575,
|
| 356 |
+
"grad_norm": 0.4761511981487274,
|
| 357 |
+
"learning_rate": 9.994279065509094e-06,
|
| 358 |
+
"loss": 0.5813,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.3493150684931507,
|
| 363 |
+
"grad_norm": 0.60807204246521,
|
| 364 |
+
"learning_rate": 9.992213708831542e-06,
|
| 365 |
+
"loss": 0.5576,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.3561643835616438,
|
| 370 |
+
"grad_norm": 0.7713252902030945,
|
| 371 |
+
"learning_rate": 9.989830958551058e-06,
|
| 372 |
+
"loss": 0.5776,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.363013698630137,
|
| 377 |
+
"grad_norm": 0.5777241587638855,
|
| 378 |
+
"learning_rate": 9.987130966157486e-06,
|
| 379 |
+
"loss": 0.5973,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.3698630136986301,
|
| 384 |
+
"grad_norm": 0.4666934907436371,
|
| 385 |
+
"learning_rate": 9.984113903310206e-06,
|
| 386 |
+
"loss": 0.5383,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.3767123287671233,
|
| 391 |
+
"grad_norm": 0.6798987984657288,
|
| 392 |
+
"learning_rate": 9.98077996182722e-06,
|
| 393 |
+
"loss": 0.5371,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.3835616438356164,
|
| 398 |
+
"grad_norm": 0.6998324394226074,
|
| 399 |
+
"learning_rate": 9.977129353672951e-06,
|
| 400 |
+
"loss": 0.5708,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.3904109589041096,
|
| 405 |
+
"grad_norm": 0.46906980872154236,
|
| 406 |
+
"learning_rate": 9.973162310944769e-06,
|
| 407 |
+
"loss": 0.5777,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.3972602739726027,
|
| 412 |
+
"grad_norm": 0.595076858997345,
|
| 413 |
+
"learning_rate": 9.968879085858234e-06,
|
| 414 |
+
"loss": 0.5366,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.4041095890410959,
|
| 419 |
+
"grad_norm": 0.6636199355125427,
|
| 420 |
+
"learning_rate": 9.964279950731066e-06,
|
| 421 |
+
"loss": 0.5822,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.410958904109589,
|
| 426 |
+
"grad_norm": 0.4243064224720001,
|
| 427 |
+
"learning_rate": 9.959365197965824e-06,
|
| 428 |
+
"loss": 0.5814,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.4178082191780822,
|
| 433 |
+
"grad_norm": 0.49584805965423584,
|
| 434 |
+
"learning_rate": 9.954135140031322e-06,
|
| 435 |
+
"loss": 0.5594,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.4246575342465753,
|
| 440 |
+
"grad_norm": 0.6090409755706787,
|
| 441 |
+
"learning_rate": 9.948590109442755e-06,
|
| 442 |
+
"loss": 0.5516,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.4315068493150685,
|
| 447 |
+
"grad_norm": 0.5092521905899048,
|
| 448 |
+
"learning_rate": 9.942730458740568e-06,
|
| 449 |
+
"loss": 0.5335,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.4383561643835616,
|
| 454 |
+
"grad_norm": 0.4809435307979584,
|
| 455 |
+
"learning_rate": 9.936556560468037e-06,
|
| 456 |
+
"loss": 0.5421,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.4452054794520548,
|
| 461 |
+
"grad_norm": 0.545276939868927,
|
| 462 |
+
"learning_rate": 9.930068807147585e-06,
|
| 463 |
+
"loss": 0.5375,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.4520547945205479,
|
| 468 |
+
"grad_norm": 0.4461447596549988,
|
| 469 |
+
"learning_rate": 9.923267611255824e-06,
|
| 470 |
+
"loss": 0.5605,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.4589041095890411,
|
| 475 |
+
"grad_norm": 0.4646362364292145,
|
| 476 |
+
"learning_rate": 9.916153405197333e-06,
|
| 477 |
+
"loss": 0.5229,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.4657534246575342,
|
| 482 |
+
"grad_norm": 0.40728962421417236,
|
| 483 |
+
"learning_rate": 9.908726641277167e-06,
|
| 484 |
+
"loss": 0.5364,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.4726027397260274,
|
| 489 |
+
"grad_norm": 0.4487316906452179,
|
| 490 |
+
"learning_rate": 9.9009877916721e-06,
|
| 491 |
+
"loss": 0.5351,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.4794520547945205,
|
| 496 |
+
"grad_norm": 0.4371180832386017,
|
| 497 |
+
"learning_rate": 9.8929373484006e-06,
|
| 498 |
+
"loss": 0.543,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.4863013698630137,
|
| 503 |
+
"grad_norm": 0.40577659010887146,
|
| 504 |
+
"learning_rate": 9.884575823291561e-06,
|
| 505 |
+
"loss": 0.5959,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.4931506849315068,
|
| 510 |
+
"grad_norm": 0.45803892612457275,
|
| 511 |
+
"learning_rate": 9.875903747951742e-06,
|
| 512 |
+
"loss": 0.569,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.5,
|
| 517 |
+
"grad_norm": 0.41057565808296204,
|
| 518 |
+
"learning_rate": 9.866921673731991e-06,
|
| 519 |
+
"loss": 0.5376,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.5068493150684932,
|
| 524 |
+
"grad_norm": 0.40417274832725525,
|
| 525 |
+
"learning_rate": 9.857630171692175e-06,
|
| 526 |
+
"loss": 0.5356,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.5136986301369864,
|
| 531 |
+
"grad_norm": 0.4480303227901459,
|
| 532 |
+
"learning_rate": 9.848029832564875e-06,
|
| 533 |
+
"loss": 0.5151,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5205479452054794,
|
| 538 |
+
"grad_norm": 0.4397015869617462,
|
| 539 |
+
"learning_rate": 9.83812126671784e-06,
|
| 540 |
+
"loss": 0.5274,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.5273972602739726,
|
| 545 |
+
"grad_norm": 0.4616395831108093,
|
| 546 |
+
"learning_rate": 9.827905104115167e-06,
|
| 547 |
+
"loss": 0.5365,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5342465753424658,
|
| 552 |
+
"grad_norm": 0.40124353766441345,
|
| 553 |
+
"learning_rate": 9.81738199427726e-06,
|
| 554 |
+
"loss": 0.5463,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.541095890410959,
|
| 559 |
+
"grad_norm": 0.4343346059322357,
|
| 560 |
+
"learning_rate": 9.80655260623953e-06,
|
| 561 |
+
"loss": 0.5346,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.547945205479452,
|
| 566 |
+
"grad_norm": 0.44946178793907166,
|
| 567 |
+
"learning_rate": 9.795417628509857e-06,
|
| 568 |
+
"loss": 0.5553,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5547945205479452,
|
| 573 |
+
"grad_norm": 0.416344553232193,
|
| 574 |
+
"learning_rate": 9.783977769024821e-06,
|
| 575 |
+
"loss": 0.5467,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5616438356164384,
|
| 580 |
+
"grad_norm": 0.41620340943336487,
|
| 581 |
+
"learning_rate": 9.772233755104695e-06,
|
| 582 |
+
"loss": 0.5365,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.5684931506849316,
|
| 587 |
+
"grad_norm": 0.43192237615585327,
|
| 588 |
+
"learning_rate": 9.76018633340719e-06,
|
| 589 |
+
"loss": 0.5366,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.5753424657534246,
|
| 594 |
+
"grad_norm": 0.4866390526294708,
|
| 595 |
+
"learning_rate": 9.747836269880005e-06,
|
| 596 |
+
"loss": 0.5549,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.5821917808219178,
|
| 601 |
+
"grad_norm": 0.5592001676559448,
|
| 602 |
+
"learning_rate": 9.73518434971211e-06,
|
| 603 |
+
"loss": 0.5605,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.589041095890411,
|
| 608 |
+
"grad_norm": 0.4245613217353821,
|
| 609 |
+
"learning_rate": 9.722231377283841e-06,
|
| 610 |
+
"loss": 0.5304,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.5958904109589042,
|
| 615 |
+
"grad_norm": 0.434048056602478,
|
| 616 |
+
"learning_rate": 9.70897817611575e-06,
|
| 617 |
+
"loss": 0.5392,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.6027397260273972,
|
| 622 |
+
"grad_norm": 0.36968135833740234,
|
| 623 |
+
"learning_rate": 9.695425588816248e-06,
|
| 624 |
+
"loss": 0.536,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.6095890410958904,
|
| 629 |
+
"grad_norm": 0.4373226463794708,
|
| 630 |
+
"learning_rate": 9.681574477028039e-06,
|
| 631 |
+
"loss": 0.531,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.6164383561643836,
|
| 636 |
+
"grad_norm": 0.3857197165489197,
|
| 637 |
+
"learning_rate": 9.667425721373333e-06,
|
| 638 |
+
"loss": 0.5239,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.6232876712328768,
|
| 643 |
+
"grad_norm": 0.4224616587162018,
|
| 644 |
+
"learning_rate": 9.65298022139786e-06,
|
| 645 |
+
"loss": 0.5462,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.6301369863013698,
|
| 650 |
+
"grad_norm": 0.4477905035018921,
|
| 651 |
+
"learning_rate": 9.638238895513687e-06,
|
| 652 |
+
"loss": 0.496,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.636986301369863,
|
| 657 |
+
"grad_norm": 0.38324639201164246,
|
| 658 |
+
"learning_rate": 9.623202680940811e-06,
|
| 659 |
+
"loss": 0.5521,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6438356164383562,
|
| 664 |
+
"grad_norm": 0.5522112250328064,
|
| 665 |
+
"learning_rate": 9.607872533647584e-06,
|
| 666 |
+
"loss": 0.5576,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6506849315068494,
|
| 671 |
+
"grad_norm": 0.45175299048423767,
|
| 672 |
+
"learning_rate": 9.592249428289935e-06,
|
| 673 |
+
"loss": 0.5461,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6575342465753424,
|
| 678 |
+
"grad_norm": 0.5615484714508057,
|
| 679 |
+
"learning_rate": 9.5763343581494e-06,
|
| 680 |
+
"loss": 0.5065,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.6643835616438356,
|
| 685 |
+
"grad_norm": 0.46789291501045227,
|
| 686 |
+
"learning_rate": 9.560128335069971e-06,
|
| 687 |
+
"loss": 0.5649,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.6712328767123288,
|
| 692 |
+
"grad_norm": 0.45602020621299744,
|
| 693 |
+
"learning_rate": 9.543632389393767e-06,
|
| 694 |
+
"loss": 0.5235,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.678082191780822,
|
| 699 |
+
"grad_norm": 0.5114532113075256,
|
| 700 |
+
"learning_rate": 9.526847569895529e-06,
|
| 701 |
+
"loss": 0.5454,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.684931506849315,
|
| 706 |
+
"grad_norm": 0.39697399735450745,
|
| 707 |
+
"learning_rate": 9.50977494371594e-06,
|
| 708 |
+
"loss": 0.5445,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.6917808219178082,
|
| 713 |
+
"grad_norm": 0.47998112440109253,
|
| 714 |
+
"learning_rate": 9.49241559629377e-06,
|
| 715 |
+
"loss": 0.5241,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.6986301369863014,
|
| 720 |
+
"grad_norm": 0.4278751313686371,
|
| 721 |
+
"learning_rate": 9.474770631296882e-06,
|
| 722 |
+
"loss": 0.512,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.7054794520547946,
|
| 727 |
+
"grad_norm": 0.5100277662277222,
|
| 728 |
+
"learning_rate": 9.456841170552054e-06,
|
| 729 |
+
"loss": 0.5387,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.7123287671232876,
|
| 734 |
+
"grad_norm": 0.41954633593559265,
|
| 735 |
+
"learning_rate": 9.438628353973654e-06,
|
| 736 |
+
"loss": 0.5454,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.7191780821917808,
|
| 741 |
+
"grad_norm": 0.42173242568969727,
|
| 742 |
+
"learning_rate": 9.420133339491171e-06,
|
| 743 |
+
"loss": 0.4957,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.726027397260274,
|
| 748 |
+
"grad_norm": 0.5260700583457947,
|
| 749 |
+
"learning_rate": 9.4013573029756e-06,
|
| 750 |
+
"loss": 0.5745,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.7328767123287672,
|
| 755 |
+
"grad_norm": 0.3974344730377197,
|
| 756 |
+
"learning_rate": 9.382301438164673e-06,
|
| 757 |
+
"loss": 0.5289,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.7397260273972602,
|
| 762 |
+
"grad_norm": 0.4914883077144623,
|
| 763 |
+
"learning_rate": 9.36296695658697e-06,
|
| 764 |
+
"loss": 0.5337,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.7465753424657534,
|
| 769 |
+
"grad_norm": 0.454052597284317,
|
| 770 |
+
"learning_rate": 9.343355087484893e-06,
|
| 771 |
+
"loss": 0.5377,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.7534246575342466,
|
| 776 |
+
"grad_norm": 0.42050087451934814,
|
| 777 |
+
"learning_rate": 9.323467077736513e-06,
|
| 778 |
+
"loss": 0.5216,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.7602739726027398,
|
| 783 |
+
"grad_norm": 0.506266713142395,
|
| 784 |
+
"learning_rate": 9.303304191776291e-06,
|
| 785 |
+
"loss": 0.502,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.7671232876712328,
|
| 790 |
+
"grad_norm": 0.44245612621307373,
|
| 791 |
+
"learning_rate": 9.282867711514703e-06,
|
| 792 |
+
"loss": 0.548,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.773972602739726,
|
| 797 |
+
"grad_norm": 0.4844461679458618,
|
| 798 |
+
"learning_rate": 9.262158936256717e-06,
|
| 799 |
+
"loss": 0.5564,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.7808219178082192,
|
| 804 |
+
"grad_norm": 0.44854727387428284,
|
| 805 |
+
"learning_rate": 9.241179182619207e-06,
|
| 806 |
+
"loss": 0.495,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.7876712328767124,
|
| 811 |
+
"grad_norm": 0.44603100419044495,
|
| 812 |
+
"learning_rate": 9.219929784447232e-06,
|
| 813 |
+
"loss": 0.5371,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.7945205479452054,
|
| 818 |
+
"grad_norm": 0.4621168076992035,
|
| 819 |
+
"learning_rate": 9.19841209272924e-06,
|
| 820 |
+
"loss": 0.5636,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.8013698630136986,
|
| 825 |
+
"grad_norm": 0.36872074007987976,
|
| 826 |
+
"learning_rate": 9.176627475511171e-06,
|
| 827 |
+
"loss": 0.5362,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.8082191780821918,
|
| 832 |
+
"grad_norm": 0.442421019077301,
|
| 833 |
+
"learning_rate": 9.154577317809483e-06,
|
| 834 |
+
"loss": 0.5426,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.815068493150685,
|
| 839 |
+
"grad_norm": 0.47269773483276367,
|
| 840 |
+
"learning_rate": 9.132263021523096e-06,
|
| 841 |
+
"loss": 0.5525,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.821917808219178,
|
| 846 |
+
"grad_norm": 0.39111027121543884,
|
| 847 |
+
"learning_rate": 9.109686005344258e-06,
|
| 848 |
+
"loss": 0.5411,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.8287671232876712,
|
| 853 |
+
"grad_norm": 0.4781222641468048,
|
| 854 |
+
"learning_rate": 9.086847704668352e-06,
|
| 855 |
+
"loss": 0.5368,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.8356164383561644,
|
| 860 |
+
"grad_norm": 0.49406808614730835,
|
| 861 |
+
"learning_rate": 9.063749571502633e-06,
|
| 862 |
+
"loss": 0.5375,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.8424657534246576,
|
| 867 |
+
"grad_norm": 0.4259958565235138,
|
| 868 |
+
"learning_rate": 9.040393074373921e-06,
|
| 869 |
+
"loss": 0.5504,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.8493150684931506,
|
| 874 |
+
"grad_norm": 0.41925352811813354,
|
| 875 |
+
"learning_rate": 9.016779698235227e-06,
|
| 876 |
+
"loss": 0.5184,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.8561643835616438,
|
| 881 |
+
"grad_norm": 0.5044152736663818,
|
| 882 |
+
"learning_rate": 8.992910944371343e-06,
|
| 883 |
+
"loss": 0.5405,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.863013698630137,
|
| 888 |
+
"grad_norm": 0.4105503261089325,
|
| 889 |
+
"learning_rate": 8.9687883303034e-06,
|
| 890 |
+
"loss": 0.5234,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.8698630136986302,
|
| 895 |
+
"grad_norm": 0.46981051564216614,
|
| 896 |
+
"learning_rate": 8.94441338969238e-06,
|
| 897 |
+
"loss": 0.5467,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.8767123287671232,
|
| 902 |
+
"grad_norm": 0.4437629282474518,
|
| 903 |
+
"learning_rate": 8.919787672241619e-06,
|
| 904 |
+
"loss": 0.5259,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.8835616438356164,
|
| 909 |
+
"grad_norm": 0.41010263562202454,
|
| 910 |
+
"learning_rate": 8.894912743598269e-06,
|
| 911 |
+
"loss": 0.557,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.8904109589041096,
|
| 916 |
+
"grad_norm": 0.4265572130680084,
|
| 917 |
+
"learning_rate": 8.869790185253766e-06,
|
| 918 |
+
"loss": 0.4936,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.8972602739726028,
|
| 923 |
+
"grad_norm": 0.516581118106842,
|
| 924 |
+
"learning_rate": 8.84442159444328e-06,
|
| 925 |
+
"loss": 0.5336,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.9041095890410958,
|
| 930 |
+
"grad_norm": 0.4106625020503998,
|
| 931 |
+
"learning_rate": 8.818808584044163e-06,
|
| 932 |
+
"loss": 0.5374,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.910958904109589,
|
| 937 |
+
"grad_norm": 0.397277295589447,
|
| 938 |
+
"learning_rate": 8.792952782473415e-06,
|
| 939 |
+
"loss": 0.5485,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.9178082191780822,
|
| 944 |
+
"grad_norm": 0.40992820262908936,
|
| 945 |
+
"learning_rate": 8.76685583358414e-06,
|
| 946 |
+
"loss": 0.5257,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.9246575342465754,
|
| 951 |
+
"grad_norm": 0.4277738332748413,
|
| 952 |
+
"learning_rate": 8.740519396561045e-06,
|
| 953 |
+
"loss": 0.5102,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.9315068493150684,
|
| 958 |
+
"grad_norm": 0.40470704436302185,
|
| 959 |
+
"learning_rate": 8.713945145814948e-06,
|
| 960 |
+
"loss": 0.5169,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.9383561643835616,
|
| 965 |
+
"grad_norm": 0.4239768385887146,
|
| 966 |
+
"learning_rate": 8.68713477087632e-06,
|
| 967 |
+
"loss": 0.5474,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.9452054794520548,
|
| 972 |
+
"grad_norm": 0.4196901023387909,
|
| 973 |
+
"learning_rate": 8.660089976287875e-06,
|
| 974 |
+
"loss": 0.511,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.952054794520548,
|
| 979 |
+
"grad_norm": 0.48978888988494873,
|
| 980 |
+
"learning_rate": 8.632812481496195e-06,
|
| 981 |
+
"loss": 0.5108,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.958904109589041,
|
| 986 |
+
"grad_norm": 0.4350135922431946,
|
| 987 |
+
"learning_rate": 8.60530402074241e-06,
|
| 988 |
+
"loss": 0.5004,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.9657534246575342,
|
| 993 |
+
"grad_norm": 0.4894917607307434,
|
| 994 |
+
"learning_rate": 8.577566342951944e-06,
|
| 995 |
+
"loss": 0.5342,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.9726027397260274,
|
| 1000 |
+
"grad_norm": 0.4722926616668701,
|
| 1001 |
+
"learning_rate": 8.549601211623316e-06,
|
| 1002 |
+
"loss": 0.5374,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.9794520547945206,
|
| 1007 |
+
"grad_norm": 0.40020692348480225,
|
| 1008 |
+
"learning_rate": 8.521410404716029e-06,
|
| 1009 |
+
"loss": 0.5241,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.9863013698630136,
|
| 1014 |
+
"grad_norm": 0.4775395691394806,
|
| 1015 |
+
"learning_rate": 8.492995714537519e-06,
|
| 1016 |
+
"loss": 0.5149,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9931506849315068,
|
| 1021 |
+
"grad_norm": 0.4094436764717102,
|
| 1022 |
+
"learning_rate": 8.46435894762922e-06,
|
| 1023 |
+
"loss": 0.5025,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.0,
|
| 1028 |
+
"grad_norm": 0.3945600688457489,
|
| 1029 |
+
"learning_rate": 8.43550192465169e-06,
|
| 1030 |
+
"loss": 0.5152,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.0068493150684932,
|
| 1035 |
+
"grad_norm": 0.48859384655952454,
|
| 1036 |
+
"learning_rate": 8.406426480268881e-06,
|
| 1037 |
+
"loss": 0.4662,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.0136986301369864,
|
| 1042 |
+
"grad_norm": 0.42010384798049927,
|
| 1043 |
+
"learning_rate": 8.377134463031468e-06,
|
| 1044 |
+
"loss": 0.4996,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.0205479452054795,
|
| 1049 |
+
"grad_norm": 0.6074568033218384,
|
| 1050 |
+
"learning_rate": 8.347627735259344e-06,
|
| 1051 |
+
"loss": 0.4944,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.0273972602739727,
|
| 1056 |
+
"grad_norm": 0.40178409218788147,
|
| 1057 |
+
"learning_rate": 8.317908172923207e-06,
|
| 1058 |
+
"loss": 0.5116,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.0342465753424657,
|
| 1063 |
+
"grad_norm": 0.4420071244239807,
|
| 1064 |
+
"learning_rate": 8.287977665525292e-06,
|
| 1065 |
+
"loss": 0.4948,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.0410958904109588,
|
| 1070 |
+
"grad_norm": 0.4358409643173218,
|
| 1071 |
+
"learning_rate": 8.257838115979244e-06,
|
| 1072 |
+
"loss": 0.4929,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.047945205479452,
|
| 1077 |
+
"grad_norm": 0.4392963945865631,
|
| 1078 |
+
"learning_rate": 8.227491440489134e-06,
|
| 1079 |
+
"loss": 0.4923,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.0547945205479452,
|
| 1084 |
+
"grad_norm": 0.44299256801605225,
|
| 1085 |
+
"learning_rate": 8.196939568427624e-06,
|
| 1086 |
+
"loss": 0.4921,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.0616438356164384,
|
| 1091 |
+
"grad_norm": 0.41032472252845764,
|
| 1092 |
+
"learning_rate": 8.166184442213314e-06,
|
| 1093 |
+
"loss": 0.4864,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.0684931506849316,
|
| 1098 |
+
"grad_norm": 0.4872773289680481,
|
| 1099 |
+
"learning_rate": 8.135228017187238e-06,
|
| 1100 |
+
"loss": 0.4849,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.0753424657534247,
|
| 1105 |
+
"grad_norm": 0.4646698236465454,
|
| 1106 |
+
"learning_rate": 8.10407226148855e-06,
|
| 1107 |
+
"loss": 0.4813,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.0821917808219177,
|
| 1112 |
+
"grad_norm": 0.42592811584472656,
|
| 1113 |
+
"learning_rate": 8.0727191559294e-06,
|
| 1114 |
+
"loss": 0.4919,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.0890410958904109,
|
| 1119 |
+
"grad_norm": 0.39574742317199707,
|
| 1120 |
+
"learning_rate": 8.041170693868985e-06,
|
| 1121 |
+
"loss": 0.4844,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.095890410958904,
|
| 1126 |
+
"grad_norm": 0.4392130970954895,
|
| 1127 |
+
"learning_rate": 8.009428881086836e-06,
|
| 1128 |
+
"loss": 0.4963,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.1027397260273972,
|
| 1133 |
+
"grad_norm": 0.47602227330207825,
|
| 1134 |
+
"learning_rate": 7.977495735655271e-06,
|
| 1135 |
+
"loss": 0.5001,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.1095890410958904,
|
| 1140 |
+
"grad_norm": 0.39845725893974304,
|
| 1141 |
+
"learning_rate": 7.945373287811116e-06,
|
| 1142 |
+
"loss": 0.4943,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.1164383561643836,
|
| 1147 |
+
"grad_norm": 0.41520991921424866,
|
| 1148 |
+
"learning_rate": 7.913063579826601e-06,
|
| 1149 |
+
"loss": 0.5106,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.1232876712328768,
|
| 1154 |
+
"grad_norm": 0.3922985792160034,
|
| 1155 |
+
"learning_rate": 7.880568665879542e-06,
|
| 1156 |
+
"loss": 0.4906,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.13013698630137,
|
| 1161 |
+
"grad_norm": 0.46458759903907776,
|
| 1162 |
+
"learning_rate": 7.847890611922721e-06,
|
| 1163 |
+
"loss": 0.4927,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.1369863013698631,
|
| 1168 |
+
"grad_norm": 0.37780556082725525,
|
| 1169 |
+
"learning_rate": 7.81503149555255e-06,
|
| 1170 |
+
"loss": 0.4639,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.143835616438356,
|
| 1175 |
+
"grad_norm": 0.3818896412849426,
|
| 1176 |
+
"learning_rate": 7.781993405876974e-06,
|
| 1177 |
+
"loss": 0.4851,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.1506849315068493,
|
| 1182 |
+
"grad_norm": 0.5063320398330688,
|
| 1183 |
+
"learning_rate": 7.748778443382658e-06,
|
| 1184 |
+
"loss": 0.4807,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.1575342465753424,
|
| 1189 |
+
"grad_norm": 0.4002656936645508,
|
| 1190 |
+
"learning_rate": 7.715388719801437e-06,
|
| 1191 |
+
"loss": 0.4678,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.1643835616438356,
|
| 1196 |
+
"grad_norm": 0.41656002402305603,
|
| 1197 |
+
"learning_rate": 7.68182635797606e-06,
|
| 1198 |
+
"loss": 0.4762,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.1712328767123288,
|
| 1203 |
+
"grad_norm": 0.42518362402915955,
|
| 1204 |
+
"learning_rate": 7.648093491725224e-06,
|
| 1205 |
+
"loss": 0.5075,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.178082191780822,
|
| 1210 |
+
"grad_norm": 0.3999499976634979,
|
| 1211 |
+
"learning_rate": 7.6141922657079045e-06,
|
| 1212 |
+
"loss": 0.4519,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.1849315068493151,
|
| 1217 |
+
"grad_norm": 0.40690121054649353,
|
| 1218 |
+
"learning_rate": 7.580124835287013e-06,
|
| 1219 |
+
"loss": 0.4662,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.191780821917808,
|
| 1224 |
+
"grad_norm": 0.4337243139743805,
|
| 1225 |
+
"learning_rate": 7.545893366392358e-06,
|
| 1226 |
+
"loss": 0.4894,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.1986301369863013,
|
| 1231 |
+
"grad_norm": 0.40752294659614563,
|
| 1232 |
+
"learning_rate": 7.511500035382943e-06,
|
| 1233 |
+
"loss": 0.4788,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.2054794520547945,
|
| 1238 |
+
"grad_norm": 0.4354107975959778,
|
| 1239 |
+
"learning_rate": 7.476947028908595e-06,
|
| 1240 |
+
"loss": 0.5209,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.2123287671232876,
|
| 1245 |
+
"grad_norm": 0.37461692094802856,
|
| 1246 |
+
"learning_rate": 7.442236543770945e-06,
|
| 1247 |
+
"loss": 0.4733,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.2191780821917808,
|
| 1252 |
+
"grad_norm": 0.3634370267391205,
|
| 1253 |
+
"learning_rate": 7.407370786783757e-06,
|
| 1254 |
+
"loss": 0.4707,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.226027397260274,
|
| 1259 |
+
"grad_norm": 0.41078001260757446,
|
| 1260 |
+
"learning_rate": 7.372351974632634e-06,
|
| 1261 |
+
"loss": 0.4897,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.2328767123287672,
|
| 1266 |
+
"grad_norm": 0.3920040428638458,
|
| 1267 |
+
"learning_rate": 7.33718233373407e-06,
|
| 1268 |
+
"loss": 0.4995,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.2397260273972603,
|
| 1273 |
+
"grad_norm": 0.400307297706604,
|
| 1274 |
+
"learning_rate": 7.3018641000939115e-06,
|
| 1275 |
+
"loss": 0.5126,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.2465753424657535,
|
| 1280 |
+
"grad_norm": 0.390546590089798,
|
| 1281 |
+
"learning_rate": 7.266399519165193e-06,
|
| 1282 |
+
"loss": 0.4777,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.2534246575342465,
|
| 1287 |
+
"grad_norm": 0.3763103187084198,
|
| 1288 |
+
"learning_rate": 7.2307908457053786e-06,
|
| 1289 |
+
"loss": 0.4714,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.2602739726027397,
|
| 1294 |
+
"grad_norm": 0.4261626899242401,
|
| 1295 |
+
"learning_rate": 7.195040343633006e-06,
|
| 1296 |
+
"loss": 0.4964,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.2671232876712328,
|
| 1301 |
+
"grad_norm": 0.38287729024887085,
|
| 1302 |
+
"learning_rate": 7.159150285883757e-06,
|
| 1303 |
+
"loss": 0.5032,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.273972602739726,
|
| 1308 |
+
"grad_norm": 0.34323829412460327,
|
| 1309 |
+
"learning_rate": 7.123122954265942e-06,
|
| 1310 |
+
"loss": 0.4533,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.2808219178082192,
|
| 1315 |
+
"grad_norm": 0.3638916313648224,
|
| 1316 |
+
"learning_rate": 7.086960639315437e-06,
|
| 1317 |
+
"loss": 0.4923,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.2876712328767124,
|
| 1322 |
+
"grad_norm": 0.41420572996139526,
|
| 1323 |
+
"learning_rate": 7.050665640150045e-06,
|
| 1324 |
+
"loss": 0.4993,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.2945205479452055,
|
| 1329 |
+
"grad_norm": 0.3671460449695587,
|
| 1330 |
+
"learning_rate": 7.0142402643233346e-06,
|
| 1331 |
+
"loss": 0.4941,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.3013698630136985,
|
| 1336 |
+
"grad_norm": 0.4007507860660553,
|
| 1337 |
+
"learning_rate": 6.977686827677926e-06,
|
| 1338 |
+
"loss": 0.5012,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.308219178082192,
|
| 1343 |
+
"grad_norm": 0.4122563302516937,
|
| 1344 |
+
"learning_rate": 6.941007654198254e-06,
|
| 1345 |
+
"loss": 0.4887,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.3150684931506849,
|
| 1350 |
+
"grad_norm": 0.4061591327190399,
|
| 1351 |
+
"learning_rate": 6.904205075862816e-06,
|
| 1352 |
+
"loss": 0.455,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.321917808219178,
|
| 1357 |
+
"grad_norm": 0.38111311197280884,
|
| 1358 |
+
"learning_rate": 6.867281432495911e-06,
|
| 1359 |
+
"loss": 0.4815,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.3287671232876712,
|
| 1364 |
+
"grad_norm": 0.41904938220977783,
|
| 1365 |
+
"learning_rate": 6.830239071618874e-06,
|
| 1366 |
+
"loss": 0.473,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.3356164383561644,
|
| 1371 |
+
"grad_norm": 0.35697585344314575,
|
| 1372 |
+
"learning_rate": 6.793080348300834e-06,
|
| 1373 |
+
"loss": 0.5002,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.3424657534246576,
|
| 1378 |
+
"grad_norm": 0.3706590533256531,
|
| 1379 |
+
"learning_rate": 6.755807625008974e-06,
|
| 1380 |
+
"loss": 0.4928,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.3493150684931507,
|
| 1385 |
+
"grad_norm": 0.37768539786338806,
|
| 1386 |
+
"learning_rate": 6.718423271458343e-06,
|
| 1387 |
+
"loss": 0.4932,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.356164383561644,
|
| 1392 |
+
"grad_norm": 0.3710521161556244,
|
| 1393 |
+
"learning_rate": 6.680929664461184e-06,
|
| 1394 |
+
"loss": 0.4774,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.3630136986301369,
|
| 1399 |
+
"grad_norm": 0.36909735202789307,
|
| 1400 |
+
"learning_rate": 6.643329187775827e-06,
|
| 1401 |
+
"loss": 0.45,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.36986301369863,
|
| 1406 |
+
"grad_norm": 0.3874911069869995,
|
| 1407 |
+
"learning_rate": 6.6056242319551315e-06,
|
| 1408 |
+
"loss": 0.4698,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.3767123287671232,
|
| 1413 |
+
"grad_norm": 0.39832133054733276,
|
| 1414 |
+
"learning_rate": 6.567817194194508e-06,
|
| 1415 |
+
"loss": 0.4626,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.3835616438356164,
|
| 1420 |
+
"grad_norm": 0.4187299311161041,
|
| 1421 |
+
"learning_rate": 6.529910478179499e-06,
|
| 1422 |
+
"loss": 0.4854,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.3904109589041096,
|
| 1427 |
+
"grad_norm": 0.3984171450138092,
|
| 1428 |
+
"learning_rate": 6.491906493932968e-06,
|
| 1429 |
+
"loss": 0.4633,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.3972602739726028,
|
| 1434 |
+
"grad_norm": 0.4085666835308075,
|
| 1435 |
+
"learning_rate": 6.45380765766187e-06,
|
| 1436 |
+
"loss": 0.477,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.404109589041096,
|
| 1441 |
+
"grad_norm": 0.4559183716773987,
|
| 1442 |
+
"learning_rate": 6.415616391603639e-06,
|
| 1443 |
+
"loss": 0.478,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.410958904109589,
|
| 1448 |
+
"grad_norm": 0.4116540551185608,
|
| 1449 |
+
"learning_rate": 6.377335123872177e-06,
|
| 1450 |
+
"loss": 0.4867,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.4178082191780823,
|
| 1455 |
+
"grad_norm": 0.34374645352363586,
|
| 1456 |
+
"learning_rate": 6.338966288303499e-06,
|
| 1457 |
+
"loss": 0.4915,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.4246575342465753,
|
| 1462 |
+
"grad_norm": 0.45456692576408386,
|
| 1463 |
+
"learning_rate": 6.300512324300975e-06,
|
| 1464 |
+
"loss": 0.4754,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.4315068493150684,
|
| 1469 |
+
"grad_norm": 0.3950651288032532,
|
| 1470 |
+
"learning_rate": 6.261975676680252e-06,
|
| 1471 |
+
"loss": 0.4742,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.4383561643835616,
|
| 1476 |
+
"grad_norm": 0.3701169192790985,
|
| 1477 |
+
"learning_rate": 6.223358795513812e-06,
|
| 1478 |
+
"loss": 0.4917,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.4452054794520548,
|
| 1483 |
+
"grad_norm": 0.45799094438552856,
|
| 1484 |
+
"learning_rate": 6.184664135975202e-06,
|
| 1485 |
+
"loss": 0.4665,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.452054794520548,
|
| 1490 |
+
"grad_norm": 0.40342605113983154,
|
| 1491 |
+
"learning_rate": 6.145894158182945e-06,
|
| 1492 |
+
"loss": 0.4827,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.4589041095890412,
|
| 1497 |
+
"grad_norm": 0.3885434567928314,
|
| 1498 |
+
"learning_rate": 6.107051327044124e-06,
|
| 1499 |
+
"loss": 0.4705,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.4657534246575343,
|
| 1504 |
+
"grad_norm": 0.41012731194496155,
|
| 1505 |
+
"learning_rate": 6.0681381120976745e-06,
|
| 1506 |
+
"loss": 0.4676,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.4726027397260273,
|
| 1511 |
+
"grad_norm": 0.40364816784858704,
|
| 1512 |
+
"learning_rate": 6.029156987357373e-06,
|
| 1513 |
+
"loss": 0.4675,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.4794520547945205,
|
| 1518 |
+
"grad_norm": 0.40886637568473816,
|
| 1519 |
+
"learning_rate": 5.990110431154549e-06,
|
| 1520 |
+
"loss": 0.4718,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.4863013698630136,
|
| 1525 |
+
"grad_norm": 0.35799840092658997,
|
| 1526 |
+
"learning_rate": 5.951000925980509e-06,
|
| 1527 |
+
"loss": 0.5037,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.4931506849315068,
|
| 1532 |
+
"grad_norm": 0.39343416690826416,
|
| 1533 |
+
"learning_rate": 5.9118309583287205e-06,
|
| 1534 |
+
"loss": 0.4883,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.5,
|
| 1539 |
+
"grad_norm": 0.4305885434150696,
|
| 1540 |
+
"learning_rate": 5.872603018536713e-06,
|
| 1541 |
+
"loss": 0.5251,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.5068493150684932,
|
| 1546 |
+
"grad_norm": 0.3737599551677704,
|
| 1547 |
+
"learning_rate": 5.8333196006277536e-06,
|
| 1548 |
+
"loss": 0.4737,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.5136986301369864,
|
| 1553 |
+
"grad_norm": 0.3764914572238922,
|
| 1554 |
+
"learning_rate": 5.793983202152283e-06,
|
| 1555 |
+
"loss": 0.508,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.5205479452054793,
|
| 1560 |
+
"grad_norm": 0.416117787361145,
|
| 1561 |
+
"learning_rate": 5.754596324029125e-06,
|
| 1562 |
+
"loss": 0.4693,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.5273972602739727,
|
| 1567 |
+
"grad_norm": 0.3617672026157379,
|
| 1568 |
+
"learning_rate": 5.715161470386485e-06,
|
| 1569 |
+
"loss": 0.4683,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.5342465753424657,
|
| 1574 |
+
"grad_norm": 0.3951134979724884,
|
| 1575 |
+
"learning_rate": 5.675681148402743e-06,
|
| 1576 |
+
"loss": 0.4884,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.541095890410959,
|
| 1581 |
+
"grad_norm": 0.3958381414413452,
|
| 1582 |
+
"learning_rate": 5.636157868147054e-06,
|
| 1583 |
+
"loss": 0.4914,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.547945205479452,
|
| 1588 |
+
"grad_norm": 0.38841086626052856,
|
| 1589 |
+
"learning_rate": 5.596594142419759e-06,
|
| 1590 |
+
"loss": 0.4879,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.5547945205479452,
|
| 1595 |
+
"grad_norm": 0.3935975730419159,
|
| 1596 |
+
"learning_rate": 5.556992486592634e-06,
|
| 1597 |
+
"loss": 0.4929,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.5616438356164384,
|
| 1602 |
+
"grad_norm": 0.4356209337711334,
|
| 1603 |
+
"learning_rate": 5.517355418448961e-06,
|
| 1604 |
+
"loss": 0.485,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.5684931506849316,
|
| 1609 |
+
"grad_norm": 0.36613333225250244,
|
| 1610 |
+
"learning_rate": 5.47768545802346e-06,
|
| 1611 |
+
"loss": 0.4623,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.5753424657534247,
|
| 1616 |
+
"grad_norm": 0.37810492515563965,
|
| 1617 |
+
"learning_rate": 5.437985127442065e-06,
|
| 1618 |
+
"loss": 0.478,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.5821917808219177,
|
| 1623 |
+
"grad_norm": 0.41265738010406494,
|
| 1624 |
+
"learning_rate": 5.398256950761578e-06,
|
| 1625 |
+
"loss": 0.508,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.589041095890411,
|
| 1630 |
+
"grad_norm": 0.3820304870605469,
|
| 1631 |
+
"learning_rate": 5.3585034538091885e-06,
|
| 1632 |
+
"loss": 0.468,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.595890410958904,
|
| 1637 |
+
"grad_norm": 0.36171847581863403,
|
| 1638 |
+
"learning_rate": 5.318727164021896e-06,
|
| 1639 |
+
"loss": 0.474,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.6027397260273972,
|
| 1644 |
+
"grad_norm": 0.4011186361312866,
|
| 1645 |
+
"learning_rate": 5.278930610285813e-06,
|
| 1646 |
+
"loss": 0.4931,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.6095890410958904,
|
| 1651 |
+
"grad_norm": 0.3681487441062927,
|
| 1652 |
+
"learning_rate": 5.239116322775392e-06,
|
| 1653 |
+
"loss": 0.464,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.6164383561643836,
|
| 1658 |
+
"grad_norm": 0.3996686339378357,
|
| 1659 |
+
"learning_rate": 5.199286832792553e-06,
|
| 1660 |
+
"loss": 0.5076,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.6232876712328768,
|
| 1665 |
+
"grad_norm": 0.40732523798942566,
|
| 1666 |
+
"learning_rate": 5.159444672605759e-06,
|
| 1667 |
+
"loss": 0.4675,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.6301369863013697,
|
| 1672 |
+
"grad_norm": 0.4019952714443207,
|
| 1673 |
+
"learning_rate": 5.119592375289015e-06,
|
| 1674 |
+
"loss": 0.4653,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.6369863013698631,
|
| 1679 |
+
"grad_norm": 0.4392927587032318,
|
| 1680 |
+
"learning_rate": 5.079732474560821e-06,
|
| 1681 |
+
"loss": 0.4886,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.643835616438356,
|
| 1686 |
+
"grad_norm": 0.42995181679725647,
|
| 1687 |
+
"learning_rate": 5.039867504623084e-06,
|
| 1688 |
+
"loss": 0.5046,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.6506849315068495,
|
| 1693 |
+
"grad_norm": 0.36546292901039124,
|
| 1694 |
+
"learning_rate": 5e-06,
|
| 1695 |
+
"loss": 0.4801,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.6575342465753424,
|
| 1700 |
+
"grad_norm": 0.4144984781742096,
|
| 1701 |
+
"learning_rate": 4.960132495376919e-06,
|
| 1702 |
+
"loss": 0.4831,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.6643835616438356,
|
| 1707 |
+
"grad_norm": 0.38707658648490906,
|
| 1708 |
+
"learning_rate": 4.92026752543918e-06,
|
| 1709 |
+
"loss": 0.4937,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.6712328767123288,
|
| 1714 |
+
"grad_norm": 0.37487655878067017,
|
| 1715 |
+
"learning_rate": 4.880407624710986e-06,
|
| 1716 |
+
"loss": 0.4648,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.678082191780822,
|
| 1721 |
+
"grad_norm": 0.38767164945602417,
|
| 1722 |
+
"learning_rate": 4.8405553273942415e-06,
|
| 1723 |
+
"loss": 0.4995,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.6849315068493151,
|
| 1728 |
+
"grad_norm": 0.401931494474411,
|
| 1729 |
+
"learning_rate": 4.800713167207449e-06,
|
| 1730 |
+
"loss": 0.4724,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.691780821917808,
|
| 1735 |
+
"grad_norm": 0.35886698961257935,
|
| 1736 |
+
"learning_rate": 4.760883677224609e-06,
|
| 1737 |
+
"loss": 0.4681,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.6986301369863015,
|
| 1742 |
+
"grad_norm": 0.3592085540294647,
|
| 1743 |
+
"learning_rate": 4.721069389714188e-06,
|
| 1744 |
+
"loss": 0.485,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.7054794520547945,
|
| 1749 |
+
"grad_norm": 0.3601101040840149,
|
| 1750 |
+
"learning_rate": 4.6812728359781064e-06,
|
| 1751 |
+
"loss": 0.4881,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.7123287671232876,
|
| 1756 |
+
"grad_norm": 0.37123242020606995,
|
| 1757 |
+
"learning_rate": 4.641496546190813e-06,
|
| 1758 |
+
"loss": 0.4868,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.7191780821917808,
|
| 1763 |
+
"grad_norm": 0.3683170974254608,
|
| 1764 |
+
"learning_rate": 4.601743049238425e-06,
|
| 1765 |
+
"loss": 0.4827,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.726027397260274,
|
| 1770 |
+
"grad_norm": 0.3582712411880493,
|
| 1771 |
+
"learning_rate": 4.562014872557936e-06,
|
| 1772 |
+
"loss": 0.4646,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.7328767123287672,
|
| 1777 |
+
"grad_norm": 0.369663268327713,
|
| 1778 |
+
"learning_rate": 4.522314541976541e-06,
|
| 1779 |
+
"loss": 0.4802,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.7397260273972601,
|
| 1784 |
+
"grad_norm": 0.3558022677898407,
|
| 1785 |
+
"learning_rate": 4.48264458155104e-06,
|
| 1786 |
+
"loss": 0.4901,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.7465753424657535,
|
| 1791 |
+
"grad_norm": 0.3494178354740143,
|
| 1792 |
+
"learning_rate": 4.443007513407368e-06,
|
| 1793 |
+
"loss": 0.4881,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.7534246575342465,
|
| 1798 |
+
"grad_norm": 0.3548129200935364,
|
| 1799 |
+
"learning_rate": 4.403405857580243e-06,
|
| 1800 |
+
"loss": 0.5079,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.7602739726027399,
|
| 1805 |
+
"grad_norm": 0.3643389940261841,
|
| 1806 |
+
"learning_rate": 4.363842131852948e-06,
|
| 1807 |
+
"loss": 0.4889,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.7671232876712328,
|
| 1812 |
+
"grad_norm": 0.3427479863166809,
|
| 1813 |
+
"learning_rate": 4.3243188515972575e-06,
|
| 1814 |
+
"loss": 0.4791,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.773972602739726,
|
| 1819 |
+
"grad_norm": 0.38511818647384644,
|
| 1820 |
+
"learning_rate": 4.2848385296135165e-06,
|
| 1821 |
+
"loss": 0.473,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.7808219178082192,
|
| 1826 |
+
"grad_norm": 0.37743064761161804,
|
| 1827 |
+
"learning_rate": 4.245403675970877e-06,
|
| 1828 |
+
"loss": 0.5084,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.7876712328767124,
|
| 1833 |
+
"grad_norm": 0.33601808547973633,
|
| 1834 |
+
"learning_rate": 4.206016797847718e-06,
|
| 1835 |
+
"loss": 0.4822,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.7945205479452055,
|
| 1840 |
+
"grad_norm": 0.3497495949268341,
|
| 1841 |
+
"learning_rate": 4.166680399372248e-06,
|
| 1842 |
+
"loss": 0.4936,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.8013698630136985,
|
| 1847 |
+
"grad_norm": 0.32923176884651184,
|
| 1848 |
+
"learning_rate": 4.127396981463289e-06,
|
| 1849 |
+
"loss": 0.4987,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.808219178082192,
|
| 1854 |
+
"grad_norm": 0.37325793504714966,
|
| 1855 |
+
"learning_rate": 4.08816904167128e-06,
|
| 1856 |
+
"loss": 0.4708,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.8150684931506849,
|
| 1861 |
+
"grad_norm": 0.3285587728023529,
|
| 1862 |
+
"learning_rate": 4.048999074019493e-06,
|
| 1863 |
+
"loss": 0.4612,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.821917808219178,
|
| 1868 |
+
"grad_norm": 0.34880074858665466,
|
| 1869 |
+
"learning_rate": 4.009889568845453e-06,
|
| 1870 |
+
"loss": 0.4805,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.8287671232876712,
|
| 1875 |
+
"grad_norm": 0.366190105676651,
|
| 1876 |
+
"learning_rate": 3.9708430126426286e-06,
|
| 1877 |
+
"loss": 0.478,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.8356164383561644,
|
| 1882 |
+
"grad_norm": 0.37551647424697876,
|
| 1883 |
+
"learning_rate": 3.9318618879023255e-06,
|
| 1884 |
+
"loss": 0.4897,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.8424657534246576,
|
| 1889 |
+
"grad_norm": 0.3399874269962311,
|
| 1890 |
+
"learning_rate": 3.8929486729558775e-06,
|
| 1891 |
+
"loss": 0.4558,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.8493150684931505,
|
| 1896 |
+
"grad_norm": 0.37752845883369446,
|
| 1897 |
+
"learning_rate": 3.854105841817056e-06,
|
| 1898 |
+
"loss": 0.476,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.856164383561644,
|
| 1903 |
+
"grad_norm": 0.38380375504493713,
|
| 1904 |
+
"learning_rate": 3.8153358640248e-06,
|
| 1905 |
+
"loss": 0.4845,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.8630136986301369,
|
| 1910 |
+
"grad_norm": 0.35891273617744446,
|
| 1911 |
+
"learning_rate": 3.776641204486191e-06,
|
| 1912 |
+
"loss": 0.4911,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.8698630136986303,
|
| 1917 |
+
"grad_norm": 0.34820806980133057,
|
| 1918 |
+
"learning_rate": 3.738024323319749e-06,
|
| 1919 |
+
"loss": 0.5001,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.8767123287671232,
|
| 1924 |
+
"grad_norm": 0.381046861410141,
|
| 1925 |
+
"learning_rate": 3.699487675699027e-06,
|
| 1926 |
+
"loss": 0.4655,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.8835616438356164,
|
| 1931 |
+
"grad_norm": 0.3240637183189392,
|
| 1932 |
+
"learning_rate": 3.661033711696501e-06,
|
| 1933 |
+
"loss": 0.4779,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.8904109589041096,
|
| 1938 |
+
"grad_norm": 0.32344701886177063,
|
| 1939 |
+
"learning_rate": 3.6226648761278238e-06,
|
| 1940 |
+
"loss": 0.4519,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.8972602739726028,
|
| 1945 |
+
"grad_norm": 0.3502044677734375,
|
| 1946 |
+
"learning_rate": 3.5843836083963625e-06,
|
| 1947 |
+
"loss": 0.4843,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.904109589041096,
|
| 1952 |
+
"grad_norm": 0.31665265560150146,
|
| 1953 |
+
"learning_rate": 3.5461923423381313e-06,
|
| 1954 |
+
"loss": 0.4679,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 1.910958904109589,
|
| 1959 |
+
"grad_norm": 0.35308602452278137,
|
| 1960 |
+
"learning_rate": 3.5080935060670345e-06,
|
| 1961 |
+
"loss": 0.4798,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.9178082191780823,
|
| 1966 |
+
"grad_norm": 0.32667192816734314,
|
| 1967 |
+
"learning_rate": 3.4700895218205026e-06,
|
| 1968 |
+
"loss": 0.4814,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.9246575342465753,
|
| 1973 |
+
"grad_norm": 0.34751468896865845,
|
| 1974 |
+
"learning_rate": 3.432182805805495e-06,
|
| 1975 |
+
"loss": 0.4553,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.9315068493150684,
|
| 1980 |
+
"grad_norm": 0.32501402497291565,
|
| 1981 |
+
"learning_rate": 3.3943757680448697e-06,
|
| 1982 |
+
"loss": 0.4725,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.9383561643835616,
|
| 1987 |
+
"grad_norm": 0.3369636535644531,
|
| 1988 |
+
"learning_rate": 3.3566708122241753e-06,
|
| 1989 |
+
"loss": 0.4462,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.9452054794520548,
|
| 1994 |
+
"grad_norm": 0.3620050549507141,
|
| 1995 |
+
"learning_rate": 3.3190703355388165e-06,
|
| 1996 |
+
"loss": 0.4758,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.952054794520548,
|
| 2001 |
+
"grad_norm": 0.33129626512527466,
|
| 2002 |
+
"learning_rate": 3.2815767285416576e-06,
|
| 2003 |
+
"loss": 0.492,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.958904109589041,
|
| 2008 |
+
"grad_norm": 0.3341040015220642,
|
| 2009 |
+
"learning_rate": 3.244192374991027e-06,
|
| 2010 |
+
"loss": 0.4611,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 1.9657534246575343,
|
| 2015 |
+
"grad_norm": 0.3279862701892853,
|
| 2016 |
+
"learning_rate": 3.2069196516991685e-06,
|
| 2017 |
+
"loss": 0.4802,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.9726027397260273,
|
| 2022 |
+
"grad_norm": 0.3518829643726349,
|
| 2023 |
+
"learning_rate": 3.169760928381127e-06,
|
| 2024 |
+
"loss": 0.4795,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.9794520547945207,
|
| 2029 |
+
"grad_norm": 0.3467582166194916,
|
| 2030 |
+
"learning_rate": 3.1327185675040907e-06,
|
| 2031 |
+
"loss": 0.4993,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.9863013698630136,
|
| 2036 |
+
"grad_norm": 0.34387102723121643,
|
| 2037 |
+
"learning_rate": 3.0957949241371845e-06,
|
| 2038 |
+
"loss": 0.4846,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 1.9931506849315068,
|
| 2043 |
+
"grad_norm": 0.3198384642601013,
|
| 2044 |
+
"learning_rate": 3.0589923458017467e-06,
|
| 2045 |
+
"loss": 0.4681,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 2.0,
|
| 2050 |
+
"grad_norm": 0.3447113633155823,
|
| 2051 |
+
"learning_rate": 3.0223131723220756e-06,
|
| 2052 |
+
"loss": 0.4804,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.006849315068493,
|
| 2057 |
+
"grad_norm": 0.3846930265426636,
|
| 2058 |
+
"learning_rate": 2.9857597356766675e-06,
|
| 2059 |
+
"loss": 0.4599,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.0136986301369864,
|
| 2064 |
+
"grad_norm": 0.3663610517978668,
|
| 2065 |
+
"learning_rate": 2.949334359849957e-06,
|
| 2066 |
+
"loss": 0.4611,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.0205479452054793,
|
| 2071 |
+
"grad_norm": 0.34175989031791687,
|
| 2072 |
+
"learning_rate": 2.913039360684565e-06,
|
| 2073 |
+
"loss": 0.4806,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.0273972602739727,
|
| 2078 |
+
"grad_norm": 0.29779112339019775,
|
| 2079 |
+
"learning_rate": 2.876877045734058e-06,
|
| 2080 |
+
"loss": 0.4257,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.0342465753424657,
|
| 2085 |
+
"grad_norm": 0.32330426573753357,
|
| 2086 |
+
"learning_rate": 2.840849714116244e-06,
|
| 2087 |
+
"loss": 0.4458,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.041095890410959,
|
| 2092 |
+
"grad_norm": 0.3471793234348297,
|
| 2093 |
+
"learning_rate": 2.8049596563669936e-06,
|
| 2094 |
+
"loss": 0.4457,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.047945205479452,
|
| 2099 |
+
"grad_norm": 0.36945298314094543,
|
| 2100 |
+
"learning_rate": 2.769209154294623e-06,
|
| 2101 |
+
"loss": 0.4343,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.0547945205479454,
|
| 2106 |
+
"grad_norm": 0.354952335357666,
|
| 2107 |
+
"learning_rate": 2.7336004808348094e-06,
|
| 2108 |
+
"loss": 0.4463,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.0616438356164384,
|
| 2113 |
+
"grad_norm": 0.3860822916030884,
|
| 2114 |
+
"learning_rate": 2.69813589990609e-06,
|
| 2115 |
+
"loss": 0.4459,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.0684931506849313,
|
| 2120 |
+
"grad_norm": 0.3559304475784302,
|
| 2121 |
+
"learning_rate": 2.662817666265932e-06,
|
| 2122 |
+
"loss": 0.4594,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.0753424657534247,
|
| 2127 |
+
"grad_norm": 0.3339487314224243,
|
| 2128 |
+
"learning_rate": 2.6276480253673663e-06,
|
| 2129 |
+
"loss": 0.4612,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.0821917808219177,
|
| 2134 |
+
"grad_norm": 0.404588520526886,
|
| 2135 |
+
"learning_rate": 2.5926292132162432e-06,
|
| 2136 |
+
"loss": 0.4606,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.089041095890411,
|
| 2141 |
+
"grad_norm": 0.3664696216583252,
|
| 2142 |
+
"learning_rate": 2.5577634562290567e-06,
|
| 2143 |
+
"loss": 0.4416,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.095890410958904,
|
| 2148 |
+
"grad_norm": 0.35626354813575745,
|
| 2149 |
+
"learning_rate": 2.5230529710914074e-06,
|
| 2150 |
+
"loss": 0.47,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.1027397260273974,
|
| 2155 |
+
"grad_norm": 0.32117801904678345,
|
| 2156 |
+
"learning_rate": 2.48849996461706e-06,
|
| 2157 |
+
"loss": 0.4475,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.1095890410958904,
|
| 2162 |
+
"grad_norm": 0.34588146209716797,
|
| 2163 |
+
"learning_rate": 2.4541066336076434e-06,
|
| 2164 |
+
"loss": 0.4436,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.1164383561643834,
|
| 2169 |
+
"grad_norm": 0.31559157371520996,
|
| 2170 |
+
"learning_rate": 2.4198751647129896e-06,
|
| 2171 |
+
"loss": 0.458,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.1232876712328768,
|
| 2176 |
+
"grad_norm": 0.3515860140323639,
|
| 2177 |
+
"learning_rate": 2.385807734292097e-06,
|
| 2178 |
+
"loss": 0.4564,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.1301369863013697,
|
| 2183 |
+
"grad_norm": 0.35864606499671936,
|
| 2184 |
+
"learning_rate": 2.3519065082747777e-06,
|
| 2185 |
+
"loss": 0.4636,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.136986301369863,
|
| 2190 |
+
"grad_norm": 0.34572356939315796,
|
| 2191 |
+
"learning_rate": 2.318173642023939e-06,
|
| 2192 |
+
"loss": 0.4431,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.143835616438356,
|
| 2197 |
+
"grad_norm": 0.33984869718551636,
|
| 2198 |
+
"learning_rate": 2.284611280198563e-06,
|
| 2199 |
+
"loss": 0.4658,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.1506849315068495,
|
| 2204 |
+
"grad_norm": 0.3217160105705261,
|
| 2205 |
+
"learning_rate": 2.251221556617344e-06,
|
| 2206 |
+
"loss": 0.4493,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.1575342465753424,
|
| 2211 |
+
"grad_norm": 0.3655526340007782,
|
| 2212 |
+
"learning_rate": 2.218006594123028e-06,
|
| 2213 |
+
"loss": 0.4517,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.1643835616438354,
|
| 2218 |
+
"grad_norm": 0.35365915298461914,
|
| 2219 |
+
"learning_rate": 2.184968504447453e-06,
|
| 2220 |
+
"loss": 0.4542,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.171232876712329,
|
| 2225 |
+
"grad_norm": 0.30709370970726013,
|
| 2226 |
+
"learning_rate": 2.15210938807728e-06,
|
| 2227 |
+
"loss": 0.4265,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.1780821917808217,
|
| 2232 |
+
"grad_norm": 0.31368520855903625,
|
| 2233 |
+
"learning_rate": 2.11943133412046e-06,
|
| 2234 |
+
"loss": 0.4563,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.184931506849315,
|
| 2239 |
+
"grad_norm": 0.3410292863845825,
|
| 2240 |
+
"learning_rate": 2.086936420173399e-06,
|
| 2241 |
+
"loss": 0.4515,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.191780821917808,
|
| 2246 |
+
"grad_norm": 0.3358498215675354,
|
| 2247 |
+
"learning_rate": 2.0546267121888863e-06,
|
| 2248 |
+
"loss": 0.4197,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.1986301369863015,
|
| 2253 |
+
"grad_norm": 0.3185282051563263,
|
| 2254 |
+
"learning_rate": 2.0225042643447283e-06,
|
| 2255 |
+
"loss": 0.4613,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.2054794520547945,
|
| 2260 |
+
"grad_norm": 0.3557981252670288,
|
| 2261 |
+
"learning_rate": 1.990571118913166e-06,
|
| 2262 |
+
"loss": 0.4824,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.212328767123288,
|
| 2267 |
+
"grad_norm": 0.31085366010665894,
|
| 2268 |
+
"learning_rate": 1.9588293061310165e-06,
|
| 2269 |
+
"loss": 0.4413,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.219178082191781,
|
| 2274 |
+
"grad_norm": 0.3326927125453949,
|
| 2275 |
+
"learning_rate": 1.9272808440706024e-06,
|
| 2276 |
+
"loss": 0.453,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.2260273972602738,
|
| 2281 |
+
"grad_norm": 0.32726767659187317,
|
| 2282 |
+
"learning_rate": 1.8959277385114516e-06,
|
| 2283 |
+
"loss": 0.454,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.232876712328767,
|
| 2288 |
+
"grad_norm": 0.3480210602283478,
|
| 2289 |
+
"learning_rate": 1.864771982812763e-06,
|
| 2290 |
+
"loss": 0.4168,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.23972602739726,
|
| 2295 |
+
"grad_norm": 0.311626672744751,
|
| 2296 |
+
"learning_rate": 1.8338155577866873e-06,
|
| 2297 |
+
"loss": 0.4566,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.2465753424657535,
|
| 2302 |
+
"grad_norm": 0.2938528060913086,
|
| 2303 |
+
"learning_rate": 1.8030604315723765e-06,
|
| 2304 |
+
"loss": 0.4372,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.2534246575342465,
|
| 2309 |
+
"grad_norm": 0.2981064021587372,
|
| 2310 |
+
"learning_rate": 1.7725085595108682e-06,
|
| 2311 |
+
"loss": 0.4396,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.26027397260274,
|
| 2316 |
+
"grad_norm": 0.3302954435348511,
|
| 2317 |
+
"learning_rate": 1.7421618840207576e-06,
|
| 2318 |
+
"loss": 0.4634,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.267123287671233,
|
| 2323 |
+
"grad_norm": 0.31649407744407654,
|
| 2324 |
+
"learning_rate": 1.71202233447471e-06,
|
| 2325 |
+
"loss": 0.4425,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.2739726027397262,
|
| 2330 |
+
"grad_norm": 0.30021733045578003,
|
| 2331 |
+
"learning_rate": 1.682091827076796e-06,
|
| 2332 |
+
"loss": 0.4365,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.280821917808219,
|
| 2337 |
+
"grad_norm": 0.30878108739852905,
|
| 2338 |
+
"learning_rate": 1.6523722647406575e-06,
|
| 2339 |
+
"loss": 0.435,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.287671232876712,
|
| 2344 |
+
"grad_norm": 0.33258000016212463,
|
| 2345 |
+
"learning_rate": 1.6228655369685342e-06,
|
| 2346 |
+
"loss": 0.4679,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.2945205479452055,
|
| 2351 |
+
"grad_norm": 0.3215084373950958,
|
| 2352 |
+
"learning_rate": 1.5935735197311204e-06,
|
| 2353 |
+
"loss": 0.4452,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.3013698630136985,
|
| 2358 |
+
"grad_norm": 0.32681819796562195,
|
| 2359 |
+
"learning_rate": 1.5644980753483109e-06,
|
| 2360 |
+
"loss": 0.4886,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.308219178082192,
|
| 2365 |
+
"grad_norm": 0.3088940382003784,
|
| 2366 |
+
"learning_rate": 1.5356410523707828e-06,
|
| 2367 |
+
"loss": 0.4603,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.315068493150685,
|
| 2372 |
+
"grad_norm": 0.3307572901248932,
|
| 2373 |
+
"learning_rate": 1.5070042854624833e-06,
|
| 2374 |
+
"loss": 0.4558,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.3219178082191783,
|
| 2379 |
+
"grad_norm": 0.31423696875572205,
|
| 2380 |
+
"learning_rate": 1.4785895952839735e-06,
|
| 2381 |
+
"loss": 0.4485,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.328767123287671,
|
| 2386 |
+
"grad_norm": 0.32906076312065125,
|
| 2387 |
+
"learning_rate": 1.4503987883766857e-06,
|
| 2388 |
+
"loss": 0.4587,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.3356164383561646,
|
| 2393 |
+
"grad_norm": 0.31784313917160034,
|
| 2394 |
+
"learning_rate": 1.4224336570480574e-06,
|
| 2395 |
+
"loss": 0.457,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.3424657534246576,
|
| 2400 |
+
"grad_norm": 0.30230167508125305,
|
| 2401 |
+
"learning_rate": 1.3946959792575915e-06,
|
| 2402 |
+
"loss": 0.4805,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.3493150684931505,
|
| 2407 |
+
"grad_norm": 0.33398470282554626,
|
| 2408 |
+
"learning_rate": 1.3671875185038064e-06,
|
| 2409 |
+
"loss": 0.4369,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.356164383561644,
|
| 2414 |
+
"grad_norm": 0.35357654094696045,
|
| 2415 |
+
"learning_rate": 1.3399100237121266e-06,
|
| 2416 |
+
"loss": 0.4409,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.363013698630137,
|
| 2421 |
+
"grad_norm": 0.344047486782074,
|
| 2422 |
+
"learning_rate": 1.312865229123681e-06,
|
| 2423 |
+
"loss": 0.4811,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 2.3698630136986303,
|
| 2428 |
+
"grad_norm": 0.30871498584747314,
|
| 2429 |
+
"learning_rate": 1.2860548541850542e-06,
|
| 2430 |
+
"loss": 0.4625,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 2.3767123287671232,
|
| 2435 |
+
"grad_norm": 0.3363014757633209,
|
| 2436 |
+
"learning_rate": 1.2594806034389556e-06,
|
| 2437 |
+
"loss": 0.4463,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 2.383561643835616,
|
| 2442 |
+
"grad_norm": 0.3206148147583008,
|
| 2443 |
+
"learning_rate": 1.233144166415861e-06,
|
| 2444 |
+
"loss": 0.4366,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.3904109589041096,
|
| 2449 |
+
"grad_norm": 0.3465731143951416,
|
| 2450 |
+
"learning_rate": 1.2070472175265857e-06,
|
| 2451 |
+
"loss": 0.4448,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.3972602739726026,
|
| 2456 |
+
"grad_norm": 0.29774489998817444,
|
| 2457 |
+
"learning_rate": 1.1811914159558374e-06,
|
| 2458 |
+
"loss": 0.4441,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.404109589041096,
|
| 2463 |
+
"grad_norm": 0.29331234097480774,
|
| 2464 |
+
"learning_rate": 1.155578405556722e-06,
|
| 2465 |
+
"loss": 0.4299,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 2.410958904109589,
|
| 2470 |
+
"grad_norm": 0.3341335952281952,
|
| 2471 |
+
"learning_rate": 1.1302098147462348e-06,
|
| 2472 |
+
"loss": 0.4509,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 2.4178082191780823,
|
| 2477 |
+
"grad_norm": 0.3187620937824249,
|
| 2478 |
+
"learning_rate": 1.1050872564017329e-06,
|
| 2479 |
+
"loss": 0.4656,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 2.4246575342465753,
|
| 2484 |
+
"grad_norm": 0.33351418375968933,
|
| 2485 |
+
"learning_rate": 1.080212327758382e-06,
|
| 2486 |
+
"loss": 0.437,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.4315068493150687,
|
| 2491 |
+
"grad_norm": 0.311325341463089,
|
| 2492 |
+
"learning_rate": 1.0555866103076212e-06,
|
| 2493 |
+
"loss": 0.4551,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 2.4383561643835616,
|
| 2498 |
+
"grad_norm": 0.30476653575897217,
|
| 2499 |
+
"learning_rate": 1.0312116696966012e-06,
|
| 2500 |
+
"loss": 0.4347,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.4452054794520546,
|
| 2505 |
+
"grad_norm": 0.3320573568344116,
|
| 2506 |
+
"learning_rate": 1.0070890556286578e-06,
|
| 2507 |
+
"loss": 0.4355,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 2.452054794520548,
|
| 2512 |
+
"grad_norm": 0.2996487021446228,
|
| 2513 |
+
"learning_rate": 9.832203017647746e-07,
|
| 2514 |
+
"loss": 0.4531,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 2.458904109589041,
|
| 2519 |
+
"grad_norm": 0.31164291501045227,
|
| 2520 |
+
"learning_rate": 9.596069256260792e-07,
|
| 2521 |
+
"loss": 0.4914,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 2.4657534246575343,
|
| 2526 |
+
"grad_norm": 0.3011796176433563,
|
| 2527 |
+
"learning_rate": 9.362504284973683e-07,
|
| 2528 |
+
"loss": 0.425,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 2.4726027397260273,
|
| 2533 |
+
"grad_norm": 0.29375994205474854,
|
| 2534 |
+
"learning_rate": 9.131522953316502e-07,
|
| 2535 |
+
"loss": 0.4695,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 2.4794520547945207,
|
| 2540 |
+
"grad_norm": 0.30826085805892944,
|
| 2541 |
+
"learning_rate": 8.903139946557437e-07,
|
| 2542 |
+
"loss": 0.457,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 2.4863013698630136,
|
| 2547 |
+
"grad_norm": 0.32142892479896545,
|
| 2548 |
+
"learning_rate": 8.677369784769041e-07,
|
| 2549 |
+
"loss": 0.4567,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 2.493150684931507,
|
| 2554 |
+
"grad_norm": 0.2887185215950012,
|
| 2555 |
+
"learning_rate": 8.454226821905171e-07,
|
| 2556 |
+
"loss": 0.4601,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 2.5,
|
| 2561 |
+
"grad_norm": 0.33354097604751587,
|
| 2562 |
+
"learning_rate": 8.233725244888291e-07,
|
| 2563 |
+
"loss": 0.4549,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 2.506849315068493,
|
| 2568 |
+
"grad_norm": 0.29360541701316833,
|
| 2569 |
+
"learning_rate": 8.015879072707611e-07,
|
| 2570 |
+
"loss": 0.4755,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 2.5136986301369864,
|
| 2575 |
+
"grad_norm": 0.30357393622398376,
|
| 2576 |
+
"learning_rate": 7.800702155527695e-07,
|
| 2577 |
+
"loss": 0.437,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 2.5205479452054793,
|
| 2582 |
+
"grad_norm": 0.2883985638618469,
|
| 2583 |
+
"learning_rate": 7.588208173807943e-07,
|
| 2584 |
+
"loss": 0.4511,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 2.5273972602739727,
|
| 2589 |
+
"grad_norm": 0.31914854049682617,
|
| 2590 |
+
"learning_rate": 7.378410637432848e-07,
|
| 2591 |
+
"loss": 0.4499,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 2.5342465753424657,
|
| 2596 |
+
"grad_norm": 0.30720996856689453,
|
| 2597 |
+
"learning_rate": 7.171322884852988e-07,
|
| 2598 |
+
"loss": 0.4815,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 2.541095890410959,
|
| 2603 |
+
"grad_norm": 0.2876569926738739,
|
| 2604 |
+
"learning_rate": 6.966958082237096e-07,
|
| 2605 |
+
"loss": 0.4205,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 2.547945205479452,
|
| 2610 |
+
"grad_norm": 0.3016420900821686,
|
| 2611 |
+
"learning_rate": 6.765329222634892e-07,
|
| 2612 |
+
"loss": 0.48,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 2.5547945205479454,
|
| 2617 |
+
"grad_norm": 0.3060811161994934,
|
| 2618 |
+
"learning_rate": 6.566449125151086e-07,
|
| 2619 |
+
"loss": 0.4609,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 2.5616438356164384,
|
| 2624 |
+
"grad_norm": 0.2952597141265869,
|
| 2625 |
+
"learning_rate": 6.370330434130317e-07,
|
| 2626 |
+
"loss": 0.4407,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 2.5684931506849313,
|
| 2631 |
+
"grad_norm": 0.3057716488838196,
|
| 2632 |
+
"learning_rate": 6.176985618353282e-07,
|
| 2633 |
+
"loss": 0.4477,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 2.5753424657534247,
|
| 2638 |
+
"grad_norm": 0.2792918086051941,
|
| 2639 |
+
"learning_rate": 5.986426970244009e-07,
|
| 2640 |
+
"loss": 0.4489,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 2.5821917808219177,
|
| 2645 |
+
"grad_norm": 0.3092077672481537,
|
| 2646 |
+
"learning_rate": 5.798666605088293e-07,
|
| 2647 |
+
"loss": 0.4356,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 2.589041095890411,
|
| 2652 |
+
"grad_norm": 0.27638155221939087,
|
| 2653 |
+
"learning_rate": 5.613716460263485e-07,
|
| 2654 |
+
"loss": 0.4553,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 2.595890410958904,
|
| 2659 |
+
"grad_norm": 0.29136520624160767,
|
| 2660 |
+
"learning_rate": 5.431588294479479e-07,
|
| 2661 |
+
"loss": 0.4275,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 2.602739726027397,
|
| 2666 |
+
"grad_norm": 0.2942065894603729,
|
| 2667 |
+
"learning_rate": 5.252293687031196e-07,
|
| 2668 |
+
"loss": 0.4159,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 2.6095890410958904,
|
| 2673 |
+
"grad_norm": 0.3121497929096222,
|
| 2674 |
+
"learning_rate": 5.075844037062322e-07,
|
| 2675 |
+
"loss": 0.4503,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 2.616438356164384,
|
| 2680 |
+
"grad_norm": 0.3024226129055023,
|
| 2681 |
+
"learning_rate": 4.902250562840622e-07,
|
| 2682 |
+
"loss": 0.4525,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 2.6232876712328768,
|
| 2687 |
+
"grad_norm": 0.30542418360710144,
|
| 2688 |
+
"learning_rate": 4.7315243010447156e-07,
|
| 2689 |
+
"loss": 0.4335,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 2.6301369863013697,
|
| 2694 |
+
"grad_norm": 0.3106946051120758,
|
| 2695 |
+
"learning_rate": 4.5636761060623314e-07,
|
| 2696 |
+
"loss": 0.4627,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 2.636986301369863,
|
| 2701 |
+
"grad_norm": 0.294289231300354,
|
| 2702 |
+
"learning_rate": 4.398716649300311e-07,
|
| 2703 |
+
"loss": 0.4503,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 2.643835616438356,
|
| 2708 |
+
"grad_norm": 0.29368045926094055,
|
| 2709 |
+
"learning_rate": 4.2366564185060134e-07,
|
| 2710 |
+
"loss": 0.4447,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 2.6506849315068495,
|
| 2715 |
+
"grad_norm": 0.2810214161872864,
|
| 2716 |
+
"learning_rate": 4.077505717100666e-07,
|
| 2717 |
+
"loss": 0.4497,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 2.6575342465753424,
|
| 2722 |
+
"grad_norm": 0.33020564913749695,
|
| 2723 |
+
"learning_rate": 3.921274663524183e-07,
|
| 2724 |
+
"loss": 0.4356,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 2.6643835616438354,
|
| 2729 |
+
"grad_norm": 0.2780897319316864,
|
| 2730 |
+
"learning_rate": 3.767973190591906e-07,
|
| 2731 |
+
"loss": 0.3993,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 2.671232876712329,
|
| 2736 |
+
"grad_norm": 0.3095431625843048,
|
| 2737 |
+
"learning_rate": 3.61761104486314e-07,
|
| 2738 |
+
"loss": 0.446,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 2.678082191780822,
|
| 2743 |
+
"grad_norm": 0.3015943765640259,
|
| 2744 |
+
"learning_rate": 3.4701977860213956e-07,
|
| 2745 |
+
"loss": 0.4296,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 2.684931506849315,
|
| 2750 |
+
"grad_norm": 0.2791678011417389,
|
| 2751 |
+
"learning_rate": 3.3257427862666894e-07,
|
| 2752 |
+
"loss": 0.442,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 2.691780821917808,
|
| 2757 |
+
"grad_norm": 0.3039053678512573,
|
| 2758 |
+
"learning_rate": 3.184255229719624e-07,
|
| 2759 |
+
"loss": 0.4256,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 2.6986301369863015,
|
| 2764 |
+
"grad_norm": 0.2986087203025818,
|
| 2765 |
+
"learning_rate": 3.045744111837529e-07,
|
| 2766 |
+
"loss": 0.4683,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 2.7054794520547945,
|
| 2771 |
+
"grad_norm": 0.30011793971061707,
|
| 2772 |
+
"learning_rate": 2.9102182388425106e-07,
|
| 2773 |
+
"loss": 0.4459,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 2.712328767123288,
|
| 2778 |
+
"grad_norm": 0.26771464943885803,
|
| 2779 |
+
"learning_rate": 2.777686227161591e-07,
|
| 2780 |
+
"loss": 0.4355,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 2.719178082191781,
|
| 2785 |
+
"grad_norm": 0.28580793738365173,
|
| 2786 |
+
"learning_rate": 2.648156502878907e-07,
|
| 2787 |
+
"loss": 0.4607,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 2.7260273972602738,
|
| 2792 |
+
"grad_norm": 0.30609655380249023,
|
| 2793 |
+
"learning_rate": 2.5216373011999697e-07,
|
| 2794 |
+
"loss": 0.4665,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 2.732876712328767,
|
| 2799 |
+
"grad_norm": 0.30261853337287903,
|
| 2800 |
+
"learning_rate": 2.3981366659281135e-07,
|
| 2801 |
+
"loss": 0.4277,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 2.73972602739726,
|
| 2806 |
+
"grad_norm": 0.2930530905723572,
|
| 2807 |
+
"learning_rate": 2.2776624489530664e-07,
|
| 2808 |
+
"loss": 0.4553,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 2.7465753424657535,
|
| 2813 |
+
"grad_norm": 0.3095199465751648,
|
| 2814 |
+
"learning_rate": 2.1602223097517915e-07,
|
| 2815 |
+
"loss": 0.426,
|
| 2816 |
+
"step": 401
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 2.7534246575342465,
|
| 2820 |
+
"grad_norm": 0.3025897741317749,
|
| 2821 |
+
"learning_rate": 2.0458237149014347e-07,
|
| 2822 |
+
"loss": 0.4385,
|
| 2823 |
+
"step": 402
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 2.76027397260274,
|
| 2827 |
+
"grad_norm": 0.27555036544799805,
|
| 2828 |
+
"learning_rate": 1.9344739376047083e-07,
|
| 2829 |
+
"loss": 0.418,
|
| 2830 |
+
"step": 403
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 2.767123287671233,
|
| 2834 |
+
"grad_norm": 0.27457016706466675,
|
| 2835 |
+
"learning_rate": 1.8261800572274001e-07,
|
| 2836 |
+
"loss": 0.4213,
|
| 2837 |
+
"step": 404
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 2.7739726027397262,
|
| 2841 |
+
"grad_norm": 0.2816792130470276,
|
| 2842 |
+
"learning_rate": 1.7209489588483396e-07,
|
| 2843 |
+
"loss": 0.4562,
|
| 2844 |
+
"step": 405
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 2.780821917808219,
|
| 2848 |
+
"grad_norm": 0.2918860614299774,
|
| 2849 |
+
"learning_rate": 1.6187873328216142e-07,
|
| 2850 |
+
"loss": 0.4405,
|
| 2851 |
+
"step": 406
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 2.787671232876712,
|
| 2855 |
+
"grad_norm": 0.30940842628479004,
|
| 2856 |
+
"learning_rate": 1.519701674351265e-07,
|
| 2857 |
+
"loss": 0.4667,
|
| 2858 |
+
"step": 407
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 2.7945205479452055,
|
| 2862 |
+
"grad_norm": 0.2933136224746704,
|
| 2863 |
+
"learning_rate": 1.4236982830782676e-07,
|
| 2864 |
+
"loss": 0.4426,
|
| 2865 |
+
"step": 408
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 2.8013698630136985,
|
| 2869 |
+
"grad_norm": 0.2901608347892761,
|
| 2870 |
+
"learning_rate": 1.3307832626800966e-07,
|
| 2871 |
+
"loss": 0.4664,
|
| 2872 |
+
"step": 409
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 2.808219178082192,
|
| 2876 |
+
"grad_norm": 0.2942031919956207,
|
| 2877 |
+
"learning_rate": 1.2409625204825802e-07,
|
| 2878 |
+
"loss": 0.4149,
|
| 2879 |
+
"step": 410
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 2.815068493150685,
|
| 2883 |
+
"grad_norm": 0.29800334572792053,
|
| 2884 |
+
"learning_rate": 1.1542417670844075e-07,
|
| 2885 |
+
"loss": 0.4464,
|
| 2886 |
+
"step": 411
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 2.821917808219178,
|
| 2890 |
+
"grad_norm": 0.27716100215911865,
|
| 2891 |
+
"learning_rate": 1.0706265159939944e-07,
|
| 2892 |
+
"loss": 0.4472,
|
| 2893 |
+
"step": 412
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 2.828767123287671,
|
| 2897 |
+
"grad_norm": 0.2767964005470276,
|
| 2898 |
+
"learning_rate": 9.901220832790104e-08,
|
| 2899 |
+
"loss": 0.4535,
|
| 2900 |
+
"step": 413
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 2.8356164383561646,
|
| 2904 |
+
"grad_norm": 0.2769363522529602,
|
| 2905 |
+
"learning_rate": 9.12733587228326e-08,
|
| 2906 |
+
"loss": 0.4219,
|
| 2907 |
+
"step": 414
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 2.8424657534246576,
|
| 2911 |
+
"grad_norm": 0.29575496912002563,
|
| 2912 |
+
"learning_rate": 8.384659480266733e-08,
|
| 2913 |
+
"loss": 0.4549,
|
| 2914 |
+
"step": 415
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 2.8493150684931505,
|
| 2918 |
+
"grad_norm": 0.2830677032470703,
|
| 2919 |
+
"learning_rate": 7.673238874417676e-08,
|
| 2920 |
+
"loss": 0.4469,
|
| 2921 |
+
"step": 416
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 2.856164383561644,
|
| 2925 |
+
"grad_norm": 0.31160372495651245,
|
| 2926 |
+
"learning_rate": 6.9931192852416e-08,
|
| 2927 |
+
"loss": 0.4427,
|
| 2928 |
+
"step": 417
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 2.863013698630137,
|
| 2932 |
+
"grad_norm": 0.28588008880615234,
|
| 2933 |
+
"learning_rate": 6.344343953196386e-08,
|
| 2934 |
+
"loss": 0.4769,
|
| 2935 |
+
"step": 418
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 2.8698630136986303,
|
| 2939 |
+
"grad_norm": 0.2827588617801666,
|
| 2940 |
+
"learning_rate": 5.726954125943318e-08,
|
| 2941 |
+
"loss": 0.439,
|
| 2942 |
+
"step": 419
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 2.8767123287671232,
|
| 2946 |
+
"grad_norm": 0.2822471857070923,
|
| 2947 |
+
"learning_rate": 5.1409890557246876e-08,
|
| 2948 |
+
"loss": 0.4169,
|
| 2949 |
+
"step": 420
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 2.883561643835616,
|
| 2953 |
+
"grad_norm": 0.2951059937477112,
|
| 2954 |
+
"learning_rate": 4.586485996867951e-08,
|
| 2955 |
+
"loss": 0.4413,
|
| 2956 |
+
"step": 421
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 2.8904109589041096,
|
| 2960 |
+
"grad_norm": 0.28772932291030884,
|
| 2961 |
+
"learning_rate": 4.063480203417625e-08,
|
| 2962 |
+
"loss": 0.4463,
|
| 2963 |
+
"step": 422
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 2.897260273972603,
|
| 2967 |
+
"grad_norm": 0.2849860191345215,
|
| 2968 |
+
"learning_rate": 3.572004926893413e-08,
|
| 2969 |
+
"loss": 0.4184,
|
| 2970 |
+
"step": 423
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 2.904109589041096,
|
| 2974 |
+
"grad_norm": 0.2915169298648834,
|
| 2975 |
+
"learning_rate": 3.1120914141766214e-08,
|
| 2976 |
+
"loss": 0.4391,
|
| 2977 |
+
"step": 424
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 2.910958904109589,
|
| 2981 |
+
"grad_norm": 0.3080134987831116,
|
| 2982 |
+
"learning_rate": 2.683768905523243e-08,
|
| 2983 |
+
"loss": 0.4579,
|
| 2984 |
+
"step": 425
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 2.9178082191780823,
|
| 2988 |
+
"grad_norm": 0.2908354103565216,
|
| 2989 |
+
"learning_rate": 2.287064632705005e-08,
|
| 2990 |
+
"loss": 0.456,
|
| 2991 |
+
"step": 426
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 2.9246575342465753,
|
| 2995 |
+
"grad_norm": 0.29851260781288147,
|
| 2996 |
+
"learning_rate": 1.9220038172780843e-08,
|
| 2997 |
+
"loss": 0.4306,
|
| 2998 |
+
"step": 427
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 2.9315068493150687,
|
| 3002 |
+
"grad_norm": 0.2807960510253906,
|
| 3003 |
+
"learning_rate": 1.588609668979446e-08,
|
| 3004 |
+
"loss": 0.4539,
|
| 3005 |
+
"step": 428
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 2.9383561643835616,
|
| 3009 |
+
"grad_norm": 0.2773742079734802,
|
| 3010 |
+
"learning_rate": 1.286903384251581e-08,
|
| 3011 |
+
"loss": 0.4576,
|
| 3012 |
+
"step": 429
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 2.9452054794520546,
|
| 3016 |
+
"grad_norm": 0.2927360534667969,
|
| 3017 |
+
"learning_rate": 1.016904144894304e-08,
|
| 3018 |
+
"loss": 0.4464,
|
| 3019 |
+
"step": 430
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 2.952054794520548,
|
| 3023 |
+
"grad_norm": 0.28278273344039917,
|
| 3024 |
+
"learning_rate": 7.78629116845786e-09,
|
| 3025 |
+
"loss": 0.471,
|
| 3026 |
+
"step": 431
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 2.958904109589041,
|
| 3030 |
+
"grad_norm": 0.2733604907989502,
|
| 3031 |
+
"learning_rate": 5.720934490907604e-09,
|
| 3032 |
+
"loss": 0.4357,
|
| 3033 |
+
"step": 432
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 2.9657534246575343,
|
| 3037 |
+
"grad_norm": 0.2910250723361969,
|
| 3038 |
+
"learning_rate": 3.973102726976819e-09,
|
| 3039 |
+
"loss": 0.4451,
|
| 3040 |
+
"step": 433
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 2.9726027397260273,
|
| 3044 |
+
"grad_norm": 0.29007551074028015,
|
| 3045 |
+
"learning_rate": 2.542906999836725e-09,
|
| 3046 |
+
"loss": 0.4503,
|
| 3047 |
+
"step": 434
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 2.9794520547945207,
|
| 3051 |
+
"grad_norm": 0.2818272113800049,
|
| 3052 |
+
"learning_rate": 1.4304382380819771e-09,
|
| 3053 |
+
"loss": 0.4465,
|
| 3054 |
+
"step": 435
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 2.9863013698630136,
|
| 3058 |
+
"grad_norm": 0.29641109704971313,
|
| 3059 |
+
"learning_rate": 6.357671699486201e-10,
|
| 3060 |
+
"loss": 0.4726,
|
| 3061 |
+
"step": 436
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 2.993150684931507,
|
| 3065 |
+
"grad_norm": 0.3044302463531494,
|
| 3066 |
+
"learning_rate": 1.5894431881657845e-10,
|
| 3067 |
+
"loss": 0.4642,
|
| 3068 |
+
"step": 437
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 3.0,
|
| 3072 |
+
"grad_norm": 0.2847689986228943,
|
| 3073 |
+
"learning_rate": 0.0,
|
| 3074 |
+
"loss": 0.4255,
|
| 3075 |
+
"step": 438
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 3.0,
|
| 3079 |
+
"step": 438,
|
| 3080 |
+
"total_flos": 666359318511616.0,
|
| 3081 |
+
"train_loss": 0.5066342979125237,
|
| 3082 |
+
"train_runtime": 9306.6951,
|
| 3083 |
+
"train_samples_per_second": 4.514,
|
| 3084 |
+
"train_steps_per_second": 0.047
|
| 3085 |
+
}
|
| 3086 |
+
],
|
| 3087 |
+
"logging_steps": 1.0,
|
| 3088 |
+
"max_steps": 438,
|
| 3089 |
+
"num_input_tokens_seen": 0,
|
| 3090 |
+
"num_train_epochs": 3,
|
| 3091 |
+
"save_steps": 100,
|
| 3092 |
+
"stateful_callbacks": {
|
| 3093 |
+
"TrainerControl": {
|
| 3094 |
+
"args": {
|
| 3095 |
+
"should_epoch_stop": false,
|
| 3096 |
+
"should_evaluate": false,
|
| 3097 |
+
"should_log": false,
|
| 3098 |
+
"should_save": true,
|
| 3099 |
+
"should_training_stop": true
|
| 3100 |
+
},
|
| 3101 |
+
"attributes": {}
|
| 3102 |
+
}
|
| 3103 |
+
},
|
| 3104 |
+
"total_flos": 666359318511616.0,
|
| 3105 |
+
"train_batch_size": 1,
|
| 3106 |
+
"trial_name": null,
|
| 3107 |
+
"trial_params": null
|
| 3108 |
+
}
|
training_loss.png
ADDED
|